In order to quantify the ecological consequences of China's recent nationwide restoration efforts, spatial explicit information on changes in forest biomass and carbon storage over the past 20 years is crucial. However, conducting long-term biomass tracking nationwide remains challenging as it requires continuous and high-resolution monitoring. This article combines multiple types of remote sensing observations with intensive field measurements, and characterizes the changes in aboveground and belowground biomass carbon (AGBC and BGBC) of Chinese forests at a spatial resolution of 1km from 2002 to 2021 through regression and machine learning methods. The most significant increase in forest biomass carbon storage is observed in the Loess Plateau, Qinling Mountains, Southwest Karst, and Southeast forests in central and southern China. Although the comprehensive use of multi-source remote sensing data provides a powerful tool for assessing changes in forest biomass carbon, further research is needed to explore the driving factors behind observed trends in woody biomass and assess the extent to which biomass benefits will translate into biodiversity, healthy and sustainable ecosystems.
Mainly obtained through field measurements and published literature records.
(1) Based on large-scale field measurements of AGBC from 2011 to 2015, a high-resolution forest aboveground biomass map of China using synthetic aperture radar was calibrated;
(2) Extend the AGBC time series to the period of 2002-2021 based on the coverage of trees and short vegetation obtained from optical remote sensing;
(3) We calibrated the AGBC time series of certain specific regions using a microwave based long-term comprehensive VOD dataset, and established a random forest model based on in situ records from published literature to draw a forest BGBC map.
In the benchmark AGBC surveying process, we multiply the in-situ AGBC data of the forest land by the forest land proportion during the field investigation period, and convert it to the average AGBC at the grid scale. Considering the overall high quality of the Chinese land use/cover dataset developed through human-computer interaction interpretation of land satellite images, and the fact that the producer accuracy (PA) and user accuracy (UA) of forest land classification in the CLCD dataset used in this study are 73% and 85%, respectively, the benchmark AGBC surveying error caused by scale conversion based on forest ground integration is generally limited. Short term changes in climate conditions may have subtle impacts on BGB, but there is currently a lack of clear knowledge about these effects. On the contrary, woody vegetation BGB is more driven by AGB (vegetation density), and the relationship between BGB and AGB is very close (R2≥ 0.85).
| # | number | name | type |
| 1 | 41991233 | National Natural Science Foundation of China |
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.NCDC.FIGSHARE.DB6534.2024
10.6084/m9.figshare.21931161.v1